[o1dschoo1]

In order to avoid condensation, is it as simple as making sure the coolant temperature doesn't drop below the ambient temperature? I'm curious because I'd like to create a shroud, or cold-box, to place around my mora, and see if I can drop temps to near-ambient during overclocking, using a standing A/C unit that I have. Though I'm not entirely sure it will prevent condensation all together, because only the mora would be placed in the cold-box, and not the rest of my bench setup. Which means fittings and cpu/gpu blocks would be exposed to higher ambient temperatures than what would be inside the cold-box, right?

Stay above ambient or liquid electrical tape the board then use art eraser pack it in and some liquid electrical tape on the back of the board. I can help you insulate it if you wanna go sub ambient.
If you are doing a cold box ambient temps would be whats inside the box.

 

[o1dschoo1]

Ok i just reread your post if you wanna run a ac unit straight into your rad it will probably drop below ambient tbh as room temps will be higher than whats coming out of the ac .. i thought you meant by cold box is running your ac into a actual sealed box to actually create a separate low ambient enclosure.. that would not condensate.

 

[o1dschoo1]

Correct me if I'm wrong but technically, temps have to stay above the dew point, not ambient.

Correct but they are pretty close depending on a few variables like humidity.
Its alot easier to just say stay above ambient while you technically could drop a few degrees lower it starts getting tricky.

 

[o1dschoo1]

Gonna followup on my last comment. Your gonna have to do the math on dew point every time you crank that setup on and getting that close to dew point id honestly insulate anyways. All its gonna take is one screw up on that math and your condensating all over your rig. While condensation isnt a big deal if you catch it right away hit it with some denatured alcohol and let it dry its a pita and lethal to your pc id you dont catch it.

 

[o1dschoo1]

Yes, the photo I grabbed from the internet didn’t have the exhaust pipe. I tried to draw the heat in a direct line to illustrate it.

Still the same principle. A/C unit will have the be placed near a window to exhaust (move) the heat out of the environment/room.

My point is… will the A/C unit cause the loop to get below ambient? I don’t think so but I’ve never tried this. Also seems insanely inefficient

It really is. If you have a spare window unit for this just pull the fan off of it and put the condensor in a ice chest filled with glycol.. phasechange water chilling is pretty efficient and the route im probably going for benching.. thats if i dont go full singlestage phase.

 

[o1dschoo1]

If the A/C unit cools the intake air down at all, won't it be below ambient? It won't cool the room below ambient in that configuration, but it will cool the exhaust below ambient.

I have seen the exhaust from an A/C unit cool down metal objects in front of it low enough to condense water when exposed to ambient air. It is pretty common for the outside of the box to condense if you trap the exhaust somewhere. The exhaust can get pretty cold in my experience.

This was exactly what i meant. And the temp of the ambient air around his rad isnt what hes worried about. The temp around the tubing in his pc is. Thats where condensation is gonna matter. If the water temp drops below dew point its gonna condensate.
@op you are gonna have to monitor water temps

 

[o1dschoo1]

You’re probably correct as I simply don’t know what is the cold temperature being exhausted from the unit. If its like 20c lower than the intake air, of course you will quickly run into issues. As the A/C is technically trying to lower the ambient of the room (intake).

Has anyone measured what temperature that air (typically) is?

Ok so i have a window unit taken apart to build a chiller. My ambient in my room is normally 70-72 with my main unit cooling the room. I put the thermostat in the vent to see how cold the air was getting and the thermostat on the unit dropped 15 degrees very quickly. I had the unit set on 65 and it went below that i couldn't exactly measure how far as i dont have a laser thermometer but the acs thermostat showed roughly 60 degrees when the compressor kicked off and it was still trying to drop.

Ik with a water chiller made with a 7k btu unit you can get water temps around -30c so id believe the air coming out is well below ambient

 

[o1dschoo1]

Thank you for all the responses!

My idea was to use the A/C unit to drop the temps as close to ambient without getting to ambient or lower. As far as efficiency, it wouldn't exactly be a priority since this wouldn't be a daily driver kind of thing. It would just be used during the 3-5 minutes of benchmarking, depending on what I'm using to benchmark at the time. And at this point, with all the good points and information posted here, I feel like I could probably just run the A/C unit so that it is blasting cold air into the intake fans on the rad, and that may allow me some amount of drop in coolant temp, but most likely wouldn't approach ambient so I wouldn't need to worry too much about condensation. So I guess the question at that point would be, would the drop in coolant temp allow me to push the processor another few hundred megs. I would need to first see how far I can push the processor without the aid of the A/C unit.

Possibly. Your gonna need a 30c temp drop on your processor roughly as voltage is gonna scale horribly depending on how close you are to the limit. Remember colder the chip better it clocks.
Honestly ambient overclocking you wont really notice too much of a difference. When you get down to -20/30c is when it starts to get interesting

 

[o1dschoo1]

Just put the whole pc in a box that is connected to your AC that blows through the box into the room, the air coming out of the AC has no chance of condensing water as the coldest point was the evaporator just before it go's out. so long as that outlet for the AC is taped up good your pc is safe!

This a chiller box is the way to go.

 

[o1dschoo1]

Just put the whole pc in a box that is connected to your AC that blows through the box into the room, the air coming out of the AC has no chance of condensing water as the coldest point was the evaporator just before it go's out. so long as that outlet for the AC is taped up good your pc is safe!

edit: to be on the safe side run the pc for awhile after the ac is turned off to warm it back up to room temps. (also when your turn off the AC you want to turn off the fan on the AC as well at least for a few min until the pc has warmed up a fair bit.)

Actually you are gonna want to slowly raise the temp in the chiller box to ambient so it doesnt condensate. Turning the ac off and leaving the pc on would flash the temperature pretty quick.

 

[o1dschoo1]

ya i was thinking about this too, might be better to just turn off the pc and ac at the same time and let it sit.

Thats the route id go honestly. If it condensates at that point theres no power going through it so its fine. Clean boards up and bam go back at it

 

[o1dschoo1]

But should that matter? so long as the pc is hot spot would that not help? note i am talking about leaving the pc in the cool box and just letting it run with no airflow. i mean the pc fans could still be spinning but not the box air it self.

Yea think about it couple hundred watta of heat with active chilling to no cool air will actually flash fairly quick. My pc raises my room temps a few degrees and that's with a ac unit in the room. Gtx 1080 with a ln2 bios on it 2100 mhz and a 5ghz 7800x pushing probably 250-300w on its on. Gotta remember heat has to go somewhere. If you could raise the room temp and water temp at the same rate to where they are equalized then sure but that all depends on fluid used. Imma assume glycol/acetone mix

 

[o1dschoo1]

I realize going subzero for overclocking is the ideal scenario, but I just don't have the money or equipment to do it. This is the next best thing for me, even if it's a little janky and doesn't net that much more overclocking headroom.

I can show you a cheap way to do it for under 200 bucks

 

[o1dschoo1]

https://forums.evga.com/DIY-low-cost-subzero-chiller-for-benchmarking-m2951571.aspx

Read this thread. I already got a ac unit apart ready to do a chiller just debating on if i wanna go chiller route or find someone to build me a phase head for cpu only cooling. Decisions decisions.
The difference at around ambient is only 100 mhz on cpu roughly. Just allow you to bump the voltage more but is still risky to hardware as the further you go below 0c the more the atoms line up and help negate electromigration

 

[o1dschoo1]

shoot, I built a single stage phase unit that did -40c on a 955BE in 2009, cost was close to a 10kbtu window unit, so cost Isnt really a major factor, heck, you could build one for a rtx3080 price nowadays. I covered the whole board in blu tack, die-electric grease in the socket, ran that way for close to two years constantly running. When I did eventually take it apart, about 15 pins were corroded off the cpu

so theres no way to stop pin rot

 

[o1dschoo1]

for anything longer than a year, i'd say no

Just for off and on bench runs tbh about a day at a time on weekends. I shouldnt have to worry bout it or pull the cpu right

 

[o1dschoo1]

The easiest way to avoid going below the dew point; always, for certain, cheaply
is evaporative cooling.

Reqd:
External rad.
Ultrasonic mist maker/s mist going onto/through rad.
Your fans will be fine thx to centrifugal force...

You need to get the humid air out the room and/or fresh/dry air in/through the rad, or it will become saturated with humidity and then evaporation stops...

It's much cheaper than throwing electricity and expensive cooling/chiller hardware at the 'below ambient cooling' challenge.

No you absolutely do.... You will absolutely never go below ambient temperature without somekind of phase change

 

[o1dschoo1]

Evaporerative cooling can get you under ambient, look at the bong cooler thread.

Not so sure if it can get you under dew point.

(Evap is phase change though…)

Well technically yes but no it wont go under dew point its crazy big inefficient and honestly not worth.

 

[o1dschoo1]

Remember the old 'Bong Cooler' days?

Here's a fellow cooling a TEC, cooling a CPU OCd to 1580 MHz!

• Idle: -14C
• Load: 4C
• Coolant 4C below ambient in reservoir.

Nuclear Tower Water Cooling - Overclockers

ED NOTE: This article is 5 pages long. Those of you running peltier cooled T-Bird rigs are probably aware that you are pushing the limits of standard water cooling methods. Many vendors claim their coolers are capable of “handling” several

www.overclockers.com

So yes; below ambient IS possible with evaporative cooling.

So:

6 mins in
Through a rad..?
(To avoid the whole: Algae build-up and no cooling when all the water's evaporated thing..?

Soz; I'm derailing tread here

I mean technically you are lowering the ambient in the bong cooler which cools pc. So it is ambient technically.

 

[o1dschoo1]

ah...
Er... no; below ambient of the room would be 'Sub Ambient' in any argument IMHO.

But point is:
With a 100% efficient Evap Cooler; you just cant ever go below the dew point as that's the point at which water vapour in the air starts condensing out.. NOT evaporating into the air...

And much cheaper than refrigeration tech and dew point sensors etc.

I really beed to stay off of here early in the am lol. Tf was i thinking.